This research addresses a new iterative geometrical noise cancellation method for closed-form camera pose estimation based on collinearity theory. We first explain how to estimate camera position and it’s orientation by employing extra nonsingular point of the edge line of the landmark’s corner through a closed-form geometrical pose estimation algorithm. Then, we propose a new iterative noise cancellation algorithm to reduce the estimation error of camera transaction matrix, which has the most portion in camera pose estimation errors. To validate our proposed method, we test it as a computer simulation. The results show that this method is efficient, accurate and robustness. But the drawback is that the camera rotation matrix, which used in noise cancellation algorithm, may not always be inverseable.

Keywords